Current guidelines recommended patent foramen ovale (PFO) occlusion as the preferred treatment for PFO-related cryptogenic stroke (CS); however, finding the causative foramen ovale remains challenging. This study aimed to identify predictors and establish a scoring system by assessing PFO morphology and stroke-related factors.
Based on a prospective multicenter registered clinical trial, we compared data mainly derived from transesophageal echocardiography (TEE) and clinical history in patients with PFO-related CS and those without CS (non-CS) with incidental PFO. Subsequently, we explored independent predictors using logistic analysis, established a scoring system based on the results, and finally evaluated the scoring system using receiver operating characteristic (ROC) analysis and internal validation.
75 patients with PFO-related CS and 147 non-CS patients were enrolled. Multivariate logistic analysis showed that the change in PFO height, large PFO, atrial septal aneurysm (ASA), and sustained right-to-left shunt (RLS) had independent relationships with CS. Based on the odds ratio value of each independent factor, a scoring system was built: change in PFO height ≥ 1.85 mm (3 points), large PFO (2 points), ASA (5 points), sustained RLS (2 points). 0–2 points correspond to low-risk PFO, 3–5 points medium-risk PFO, and 7–12 points high-risk PFO. ROC analysis showed an area under the curve of 0.80 to predict CS. The proportion of patients with CS is increasing based on these points.
Our study screened out the change in PFO height as an independent predictor of CS. A simple and convenient scoring system can provide constructive guidance for identifying whether the PFO is causal and consequently selecting patients more likely to benefit from closure.